General battery assistance for 52v setup

[Panderpski]

Hi all, I've recently jumped into the deep end of custom ebikes and gotten to the point of making a battery. I've decided on a CYC motor/controller and want to run a couple small 52v packs. I know I want to run a config that exceed the requirements for longevity but also compact so I landed on P50Bs in a 14s2p config paired with a 100A BMS.

Questions:

1. I lack knowledge about how current works. The motor will want 90-120A (30-40A per phase?). Does this mean the pack will only need to supply 30-40A or does the pack and BMS need to be able to output the peak 120A? I'm assuming I made the correct decisions for parts since CYC and Electrify sell similar spec packs that are compatible with the motor/controller.
2. I want to keep costs down so I think I want to stack nickel as needed. What size strips would be recommended?
   • I see that 10x0.15mm Nickel should do ~6A optimally. Would that mean that every single bridge would require 10 strips to be stacked to achieve 60A, or because of the configuration would I be able to run 5 strips since each parallel would pull 30A? I could also see a case that I'd only need 1 strip since each of the 28 cells might only be outputting ~2A.
   • Keep in mind I'd like to minimize heat and maximize longevity of the packs.
3. Edit: Another question I can't seem to find the answer for: I've seen that a couple threads have stated its the connection between the parallels that need to be stacked. Does this mean that series connections do not need to be stacked? Not that it really has any bearing on this build since its 2p.

I'll probably have follow-up questions. Please correct me on anything I might not be understanding - I am a student of the electrical arts. Please be patient =)

Here's the configuration I've diagrammed in Figma in case it helps.

 

[E-HP]

Yoiu might want to be more conservative rating the BMS for your 2P pack. The cells may be capable initially, but you should choose the BMS anticipating cell degradation over time.
The battery should be capable, with some headroom, of handling the battery current the controller is set for, or its continuous rating of the controller if it’s not configurable. Phase current isn’t a third of battery current. By contrast, phase current can be 2 to 3 times battery current; but regardless, you only need to be concerned with battery (rated) current.

 

[Panderpski]

Yoiu might want to be more conservative rating the BMS for your 2P pack. The cells may be capable initially, but you should choose the BMS anticipating cell degradation over time.
The battery should be capable, with some headroom, of handling the battery current the controller is set for, or its continuous rating of the controller if it’s not configurable. Phase current isn’t a third of battery current. By contrast, phase current can be 2 to 3 times battery current; but regardless, you only need to be concerned with battery (rated) current.

How conservative should I be? I thought 100A continuous would be ample headroom looking at comparable specs.
I was taking a look at the CYC A-52 batteries (compatible with their controllers and motors), and while yes they use cells rated at much lower discharge (Samsung 50S in 14s2/3p I believe), they use a BMS rated for 50/70A continuous respectively. In their configuration, they are at the limit of what the battery pack can do. I believe Electrify Bikes also has a very similar configuration. From my research, I found that the P50B also lives longer than the Samsung cell in the other packs.

What happens to Amps as the cells degrade that would necessitate even more headroom? I was under the assumption that voltage and amps went down as cells died.

 

[docw009]

I've found that my older packs still charge up close to 4,2V/cell with a balance BMS, but I have no way to determine how much the peak current goes down. It probably decreases with age. For sure, the capacity does drop and I do measure that. I rarely, if ever, run my batteries close to the peak current of the controllers other than a few occasional throttle blips, and usually stop before 50% SOC so neither is a show stopper for me.

In your case, it just feels like 14S-2P is too small for a CVC rig, I built a 13S-1P with Molicel P42A and also one with Samsung 50S, which is like 45A and 25A peak respectively. Sure, at full charge, they will push 26A into my BBS02B, but the voltage sag is at least 4 volts. Probably can't do that much under 46V or the batteries will shut off or controller shuts off. They were measured at 205WH and 180WH respectively too, so not a lot of range. WHen riding with my wife, 180WH is good enough for 20 miles, and they were built as small spares to be used on each bike.

I also used two layers of 8mm x .15mm nickel, which I believe will support 20A surges, but run all say below 10A. I did 3 layers once, but it was unwieldy and early in my spot weld learning process. It came apart with a pen knife, Far easier to build low capacity batteries, You set yourself a challenge,

 

[Panderpski]

So a bit of a conclusion to this thread.

I ended up building the first of two batteries about 90% of the way with the structure show in the original post. Once I attempted to solder wires together, I found out that I suck at it. I had a realization the following day that I might be able to place the batteries in a different orientation that would remove the need for soldering all together. So that's what I did - I completely removed all the welds and residual material from the pack and ended up puncturing one of the cells in the process. That made me have to order a replacement. So while I waited on that replacement I built the second back with the other 28 of 56 cells I had on hand.

I went with a configuration of 2-layers of Copper/Nickel sandwich of 8x0.15mm nickel and 10x0.2mm copper. My spot welder, the AwithZ UF20B was able to do this at 90/99 gear and 2 consecutive welds. Not bad for a $150 Canadian dollar welder! This configuration easily allows for over 60 amps continuous with spike above 80 amps, though I've not seen it get even close to 60 amps.

The battery: The pack shown below is the battery composed of cells I had broken down to redo but allowed me to take learnings from the other battery and build this one better.

- I started off by arranging the cells as such where the total positive is the top left and the total negative is the bottom left in the picture below.
- Once I had them placed, I wanted to keep them tight and in the proper shape. I didn't have the right size plastic cell organizers so I placed tape on one side, then flipped the pack and wrapped 10mm Kapton tape (high-temp masking tape) around the parameter of the pack, pulling tight to allow the tape to wrap over and grab the top edge of the cell wrapper. After about 4 or 5 loops, the pack was very ridged and I could remove the tape on the opposite side that was originally used to keep the loosely pack in the right shape.
- Because I didn't have the insulation rings that are normally placed on the positive side of each cell, I ran 5mm Kapton tape over the edges of all cells in 3 directions - horizontal and two diagonals. If you look closely at the picture below, you'll see that there are two strips for each direction. Both are slightly off center. This is because I ran the tape right up to the button of the positive cell. This solution has worked very well for me.



After taping one side, I wrapped the widest Kapton tape I had on hand around the middle of the pack a few times and continued to repeat the process outlined above for the other side for the opposite side of the pack.



This is the non-total side of the pack. I did make a mistake here and had to redo a small part.
IF YOU WANT TO FOLLOW THIS AS A GUIDE: Note that the section that looks like a T should not be configured like this. The 4 cells that look like they would make the top of the T should be connected and the other four cells that make the bottom line of the T should be connected.



I then proceeded to do the opposite side, wired the BMS and charge/discharge cable. I did not run a separate balancing module since the Daly Smart Active Balancer I picked up has a 1A active balancer and works extremely well so far - I've yet to see the cells more than a 0.008v difference from the lowest and highest cell. This allows me to use one connection for both charge and discharge.

Once the battery was wired up, I did a final test by checking the total voltage with a multimeter. Since the pack was in the range of voltage I would expect (the pack was made of cells that were at about 3.5v each when I received them so the total voltage was around 49v), I fully wrapped the back in Kapton tape, green insulation paper, and finally heat shrink. I didn't cut away too much heat shrink and instead just folded it where it was bunching up while it was still hot. The shape didn't have to be perfect since the case it's in is slightly larger than the pack.



Also built a little cable to reach from the saddle mounted battery to reach the controller.



And the final look with both packs on the custom AliExpress bike I built.



I've been able to ride over 150Km in the last week and boy does it ride well. The CYC X1 Stealth Gen 3 lays down power and the output is very smooth after a little tweaking in the CYC app. Each of the battery packs give me well over 30Km. I've not been able to successfully drain one of them to the minimum I've set up in the Daly app but I think I could do 40Km with how I ride or even further if I ride like a normal person. With both packs on, I'll be adding a parallel module to allow both packs to work at the same time so I don't have to swap the cable mid-ride.